Fix HUD visibility API and showhud/player map state conflicts

[QueryOfficial]

Summary

This PR fixes two related HUD issues discovered while resolving #5026:

bool isPlayerHudComponentVisible(string component [, bool checkEnabled = true])

• checkEnabled = true (default): unchanged behavior; reports the state requested through setPlayerHudComponentVisible, so no existing script breaks and scripts can still distinguish "disabled by script" from "hidden by showhud".
• checkEnabled = false: reports the effective on-screen visibility, accounting for the global HUD state (showhud command, fullscreen player map) and every containing component in the draw path (radar_map, radar_blips and radar_altimeter inherit the radar's visibility).

Under the hood this introduces owner-aware HUD visibility state:

• The persistent visibility requested through showhud is kept separate from temporary suppression, so showhud and the fullscreen player map no longer overwrite each other's state. Closing the map can no longer re-enable a HUD previously hidden with showhud 0.
• The showhud toggle now flips only the user's own preference, so it behaves predictably while the map is open.
• Game reset clears persistent state, temporary suppression and closes the player map, so no stale suppression or chat visibility state survives a reconnect.
• GTA HUD draw instructions are only rewritten when the effective state actually changes, and SetComponentVisible skips the unprotect/write cycle when the component is already in the requested state.
• New CHud virtual methods are added at the end of the interface to preserve existing virtual method positions across client modules.

There are no Lua function signature breaks, component ID, network protocol, or migration changes. The OOP variant (player.isHudComponentVisible) supports the new parameter automatically.

Motivation

Fixes #5026.
Fixes #547.

isPlayerHudComponentVisible("radar") returned true after showhud hid the HUD, because it only inspected the per-component state set by setPlayerHudComponentVisible and ignored the global HUD draw state.

This problem has a long history, and previous attempts were closed specifically because they changed the default return value:

• isPlayerHudComponentVisible only works with setPlayerHudComponentVisible #547 — the original report. In the discussion, @qaisjp concluded that the default behavior cannot simply be changed, and proposed exactly this syntax: bool isPlayerHudComponentVisible(string component, bool checkEnabled = true), with the default documented as checking the enabled state.
• Fix isPlayerHudComponentVisible incompatible with command showhud #1609 — closed. Review feedback: making the function return the effective state by default "would make it impossible to know whether a component has been disabled by script or user", which "is a negative side effect that must be solved before merge".
• Fixed isPlayerHudComponentVisible not working with showhud #3634 — closed. @TheNormalnij: "I'd like to keep isPlayerHudComponentVisible only for setPlayerHudComponentVisible", referring back to the checkEnabled proposal from isPlayerHudComponentVisible only works with setPlayerHudComponentVisible #547 as a valid variant.

This PR implements the solution the maintainers converged on in those discussions: the default stays fully compatible, and the effective visibility is available on request via checkEnabled = false.

While fixing the API, a second issue was found and fixed: showhud and the fullscreen player map both called Disable(bool) on a single global flag, so for example after showhud 0, opening and closing the F11 map unintentionally re-enabled the HUD. Persistent requested visibility and temporary suppression are now tracked separately, and each owner releases only its own suppression.

Test plan

Build:

• Built Game SA (Release|Win32) successfully.
• Compiled all touched Client Core / Client Deathmatch translation units successfully.
• Ran utils/clang-format.ps1 after the C++ changes.

API behavior:

1. isPlayerHudComponentVisible("radar") → true.
2. showhud → isPlayerHudComponentVisible("radar") → still true (documented compatible default), isPlayerHudComponentVisible("radar", false) → false.
3. showhud again → both return true.
4. setPlayerHudComponentVisible("radar", false) → both variants return false.
5. Hide parent radar only → "radar_map", "radar_blips", "radar_altimeter" with checkEnabled = false all return false; restoring radar restores them, while an explicitly hidden radar_blips stays false until restored.

State ownership:

1. showhud 0 → HUD hidden.
2. Open F11 map → HUD still hidden; close it → HUD still hidden (showhud state preserved).
3. showhud (toggle) → HUD visible again.
4. Open F11 map with HUD visible → HUD suppressed; showhud toggle while the map is open flips only the user preference.
5. Reconnect with HUD hidden and map open → reset restores the HUD and closes the map with no stale suppression.

Checklist

• Your code should follow the coding guidelines.
• Smaller pull requests are easier to review. If your pull request is beefy, your pull request should be reviewable commit-by-commit.

[TheCrazy17]

You could also add Fixes #547 so both issues get closed by the same pull request. 😄

[QueryOfficial]

You could also add Fixes #547 so both issues get closed by the same pull request. 😄

Thanks @crazy17 ! Good point — I've updated the PR description to also include "Fixes #547", so both issues get closed by this pull request. 🎉

[FileEX]

Conflict with #5028

[QueryOfficial]

Thanks for the review @FileEX — applied your feedback:

• IsDisabled() is now const noexcept
• Replaced memcmp with std::memcmp
• Switched member/default init to brace initialization (parent{HUD_ALL}, m_isEnabled{true}, m_isAppliedDisabled{false})
• Removed the eHudSuppressionReason enum — agreed it wasn’t needed. Temporary map suppression is now a single bool via SetSuppressed(bool)
• Dropped the virtual-method-order comment